Latent Semantic Analysis
A classical method in information retrieval that uses matrix factorization to map terms and documents into a low-dimensional continuous semantic space, so that semantically similar items lie close to one another. It is the origin of the concept of the latent semantic space.
Timeline (last 90 days)
The concept of latent semantic space originates from Latent Semantic Analysis (LSA) in the information retrieval field, which maps words and documents into a low-dimensional continuous space through matrix decomposition.
SVD is used for matrix completion in recommender systems and for latent semantic analysis in natural language processing.
潜在语义分析(LSA)诞生于1988年,通过对词-文档矩阵(通常经TF-IDF加权)进行SVD并保留前k个奇异值,将词和文档映射到k维语义空间
All Facts (3)
The concept of latent semantic space originates from Latent Semantic Analysis (LSA) in the information retrieval field, which maps words and documents into a low-dimensional continuous space through matrix decomposition.
50%UnverifiedSVD is used for matrix completion in recommender systems and for latent semantic analysis in natural language processing.
50%Unverified潜在语义分析(LSA)诞生于1988年,通过对词-文档矩阵(通常经TF-IDF加权)进行SVD并保留前k个奇异值,将词和文档映射到k维语义空间
50%